help with Ms-20 filter emulation

    May 13 2012 | 2:13 am
    I've been trying to implement the MS-20 filter from this paper highlighted by Peter McCullough in a previous post about the MS-20 filter. Maybe someone has done it already I don't know. Aaaaanyhoooo my knowledge of these things is not great and I am stumped by the saturation part of the filter. Basically I've done something very wrong as the filter overloads pretty easy and resonance isn't in the correct place.
    Anyone offering help will be thanked verily.

    • May 13 2012 | 7:57 pm
      bump bee bump bump bump
    • May 14 2012 | 8:45 pm
      I've implemented it, but I haven't had great results with it as yet, as I feel the tan tracking of the frequency is pretty ineffective. For the saturation, I don't use the second clip. Also, your feedback needs to be inverted as I read it. (e.g. - at the top instead of +)
      I've had better luck with some correction factors that I've added in, and by also upsampling the filter (the response above SR/4 is terrible otherwise), but am trying to get it together into one comprehensive piece.
    • May 16 2012 | 11:37 pm
      Hey thanks peter can't believe I didn't read that [-] in the diagram...very much still a noob. I've tried up sampling the filter via poly~ too which is a definitely better response. At this point I'm kind of just messing around with it so doubt I will end up with an MS-20 filter as its more of a learning experience than anything else. I have been cascading the filter and also implemented the multiple modes lowpass, high pass, bandpass. I found this interesting when you cascade them - high pass into lowpass vs lowpass into high pass vs bandpass into bandpass.
      If you end up posting yours I'll be intrigued to look at it.